Temporal issues in person–organization fit,person–job fit and turnover: The role of leader

Abstract

Person–environment fit has been found to have significant implications for employee attitudes and behaviors. Most research to date has approached person–environment fit as a static phenomenon, and without examining how different types of person–environment fit may affect each other. In particular, little is known about the conditions under which fit with one aspect of the environment influences another aspect, as well as subsequent behavior. To address this gap we examine the role of leader–member exchange in the relationship between two types of person–environment fit over time: person–organization and person–job fit, and subsequent turnover. Using data from two waves (T1 and T2, respectively) and turnover data collected two years later (T3) from a sample of 160 employees working in an elderly care organization in the Netherlands, we find that person–organization fit at T1 is positively associated with person–job fit at T2, but only for employees in high-quality leader–member exchange relationships. Higher needs–supplies fit at T2 is associated with lower turnover at T3. In contrast, among employees in high-quality leader–member exchange relationships, the demands–abilities dimension of person–job fit at T2 is associated with higher turnover at T3.

Keywords

leader–member exchange person–environment fit person–job fit person–organization fit turnover

Research on person–environment (PE) fit has shown that congruence between the attributes of a person and the attributes of his or her environment is positively related to several desirable outcomes for employees and organizations, such as greater job satisfaction, stronger commitment, higher engagement in in-role and extra-role behaviors, and lower turnover (Hoffman and Woehr, 2006; Kristof-Brown et al., 2005; Verquer et al., 2003). Such findings hold both for actual levels of fit (i.e. based on the matching of organizational and personal values or characteristics) and for perceptions of fit (i.e. employees’ own estimations of their PE fit), with the latter seemingly serving as better predictors of employees’ attitudes and behaviors (see Kristof-Brown et al., 2005).

As PE fit is a broad concept that inherently involves one’s compatibility with multiple aspects of the work environment (Kristof-Brown et al., 2002), researchers have called for studies that consider multiple dimensions of fit, rather than fit with a single dimension (e.g. Kristof-Brown et al., 2005; Maden and Kabasakal, 2014; Tak, 2011). Indeed, results suggest that different types of fit have different effects on work-related outcomes (e.g. Lauver and Kristof-Brown, 2001). Two types of PE fit that have been shown to be important in each phase of an employee’s work experience, from entry to long-term employment, are person–organization (PO) and person–job (PJ) fit, which represent, respectively, the match between the individual and the organization, and the match between the individual and the job (Jansen and Kristof-Brown, 2006; Kristof-Brown et al., 2005).

Employees who experience poor fit with their work environment are likely to leave the organization (e.g. Schneider et al., 1995). The Theory of Work Adjustment (TWA; Dawis and Lofquist, 1984) has been posited as a useful framework for understanding the association between PE fit and turnover. The TWA describes an ongoing process of interaction (work adjustment) between workers and their work environment, suggesting that employees and organizations interact and meet each other’s respective requirements. The success of the work adjustment process is reflected in the employee’s satisfactoriness on the one hand (i.e. the extent to which the individual can meet job demands), and in his or her satisfaction from the work on the other, and is expected to influence the employee’s likelihood of remaining with the organization (e.g. Dahling and Librizzi, 2015; Hesketh, 1993). Notably, despite the conceptualization of work adjustment as an ongoing process, most prior research exploring the PE fit–turnover association adopts a static view on fit, that is, it assumes that once a specific type of fit has been achieved or, on the contrary, has become problematic, it tends to stay so for a long period. As Gabriel et al. (2014: 390) recently noted: “Almost all of the research on fit perceptions has been conducted at the between-person level of analysis, which implicitly ignores the possibility of substantive within-person changes.” Indeed, several researchers (e.g. Feldman and Vogel, 2009; Jansen and Shipp, 2013; Tinsley, 2000) argue that measuring fit at one particular moment provides an inaccurate understanding of how individuals and environments mesh in the long run. In particular, recognizing that jobs, organizations and individuals change over time and that changes in work environments create changes in individuals (e.g. massive layoffs lead to job insecurity), and vice versa (e.g. generation Y employees demand more work–life balance), these researchers point out that PE fit should be viewed as a process.

To address this lacuna, an emerging body of research has begun to conceptually consider fit as something that develops over time (Jansen and Kristof-Brown, 2006; Shipp and Jansen, 2011; Yu, 2009, 2013). Empirical evidence regarding the non-static nature of fit has also begun to accumulate. In particular, Cable and Parsons (2001) examined the degree to which different socialization tactics predict changes (over time) in one type of fit (congruence between individuals’ values and organizational values) among newcomers. Similarly, Cooper-Thomas et al. (2004) examined the effect of socialization tactics on changes in newcomers’ perceived and actual PO fit, and also investigated relations between these two measures of PO fit. Another study (Gabriel et al., 2014), has examined how variability in PO and PJ fit perceptions over time relates to work affect and job satisfaction. Finally, DeRue and Morgeson (2007) examined how person–team fit and PJ fit operate over time in team contexts, and found that the former is generally stable over time, whereas the latter changes over time.

Less is known about the degree to which different types of fit influence each other over time. A greater understanding of these effects not only adds to the growing stream of research that focuses on within-individual variability in fit perceptions, but also advances fit theory because it sets forth the proposition that fit with a single, specific work dimension should not be studied in isolation from, but rather as a potential contributor to developments in, other fit dimensions. Therefore, it is important to consider the degree to which PO fit is related to PJ fit over time. Notably, because fit with one aspect of the environment is known to spill over into other areas, a specific fit dimension may trigger a change in other fit dimensions over time (Jansen and Kristof-Brown, 2006). Along these lines, Tak (2011) has suggested that PO fit and PJ fit may gradually influence each other over time, and Shipp and Jansen (2011) suggested that multiple perceptions of PO fit and PJ fit may coexist within a person’s overall fit process.

Building on this, the main aim of this study has to do with the conditions under which perceptions of fit develop, and affect employee behavior (Erdogan et al., 2004; Kristof-Brown et al., 2005; Resick et al., 2007). Employee–supervisor relationships, which have a key role in employees’ overall work experience (e.g. Biron, 2010; Uhl-Bien and Carsten, 2007), may be of particular relevance to employees’ perceptions of fit and consequent behavior over time (Kristof-Brown et al., 2005). In particular, a good relationship with one’s supervisor may strengthen the relationship between different types of fit over time. Moreover, evidence suggests that the nature of employees’ relationships with supervisors can influence associations between fit and various organizational outcomes (e.g. Erdogan et al., 2004). In line with these arguments, we first propose that leader–member exchange (LMX) has a moderating role in the relationship between PO and PJ fit over time. Doing so, we aim to add to the TWA by showing which factors affect the ongoing process of work adjustment, as well as the success of the work adjustment process by better achieving both satisfactoriness and satisfaction. Second, we will examine the moderating role of LMX in the relationship between perceived fit and subsequent outcomes. Employee turnover, a well-documented concern for many employers that imposes high costs on both organizations and employees (e.g. Holtom et al., 2008), will be the focal outcome variable examined in this study.

To test our model, we use survey data collected in two waves (T1 and T2) and subsequent turnover data obtained two years later (T3) from a sample of 160 employees working in an elderly care organization in the Netherlands. The research model is presented in Figure 1.

Figure 1.

Research model.

The temporal relationship between person–organization and person–job fit

The merits of fit between individuals and their work environments have been studied extensively, with research in this field focusing on a variety of dimensions of PE fit, including fit with the organization, job, supervisor and work group (Kristof, 1996; Kristof-Brown et al., 2005). Analyses of PE fit typically conceptualize the construct in one of two ways: as supplementary fit or as complementary fit (Cable and Edwards, 2004; Kristof-Brown et al., 2005; Muchinsky and Monahan, 1987). Supplementary fit refers to a case in which a person’s characteristics match characteristics of the environment; this conceptualization of fit corresponds to the value congruence or similarity attraction paradigm (Cable and Edwards, 2004; Ostroff and Judge, 2007). Complementary fit occurs “when a person’s characteristics ‘make whole’ the environment or add to it what is missing” (Kristof, 1996: 3; Muchinsky and Monahan, 1987), and is based on the psychological need fulfillment paradigm (Edwards, 1991). In the current study, we focus on the two types of fit that best capture these two main conceptualizations of PE fit, namely, PO fit and PJ fit, respectively. Supplementary fit has dominated the PO fit literature, and complementary fit has been the main focus of PJ fit research (Kristof-Brown et al., 2005).

PO fit is typically conceptualized as value congruence, or the match between an employee’s values and organizational values (e.g. Cable and DeRue, 2002; Elfenbein and O’Reilly, 2007; Kristof-Brown et al., 2005). According to the similarity attraction paradigm underlying value congruence, people are more attracted to and trusting of others who are similar to them (Cable and Edwards, 2004). Prior research has accordingly demonstrated that value congruence between an organization and its employees fosters communication between employees, enhances employees’ identification with the organization, creates an environment of trust, and translates into positive work-related attitudes and behaviors (Edwards and Cable, 2009; Kristof-Brown et al., 2005; Verquer et al., 2003).

PJ fit is defined as the match between characteristics of the person and those of the job or task (Edwards, 1991; Kristof, 1996). Employees have been found to be healthier (in terms of well-being) and more successful (in terms of performance) in jobs that better fit their personal attributes (Caldwell and O’Reilly, 1990; Kristof-Brown et al., 2005; Park et al., 2011). PJ fit comprises two key components (Cable and DeRue, 2002; Edwards, 1991). The first is demands–abilities (DA) fit, which refers to the degree to which an employee’s knowledge, skills and abilities match the requirements of the job. Abilities are often measured according to aptitudes, experience and education, and job demands often take the form of work load, performance requirements and instrumental activities. The second component, needs–supplies (NS) fit, refers to the degree to which the employee’s needs are addressed by the supplies that emanate from his or her job. Employees’ needs consist of psychological desires, values, goals, interests and preferences, whereas job supplies include pay, benefits, training, interesting and challenging work, promotion opportunities, recognition, good working conditions and decision-making latitude (Cable and DeRue, 2002; Muchinsky and Monahan, 1987). Whereas DA fit focuses more on the job (job demands are fulfilled by employees), NS fit has a stronger employee focus (employee needs are fulfilled by the job).

According to the psychological need fulfillment paradigm underlying complementary fit in general and PJ fit in particular (Edwards, 1991), employees express more positive attitudes at work when their needs are more fully addressed (Locke, 1976). Accordingly, and as further suggested in the TWA (Dawis and Lofquist, 1984), employees seek to achieve and maintain correspondence with their work environment, both by fulfilling the requirements of the job (achieving DA fit, or satisfactoriness in TWA terms) and by reaching a state in which the work environment fulfills the requirements of the individual (NS fit or satisfaction in TWA terms – the extent to which the job meets the expectations and needs of the individual) (Dawis and Lofquist, 1984; Rounds et al., 1987).

As discussed above, several researchers have noted the need for further research on perceptions of fit with multiple aspects of the work environment over time (e.g. Gabriel et al., 2014; Kristof-Brown et al., 2005; Maden and Kabasakal, 2014; Shipp and Jansen, 2011; Tak, 2011). Addressing this call, we propose that the relationship between PO and PJ fit over time varies as a function of employee relationships with their supervisor; that is, LMX is offered as an enabling condition in the relationship between perceptions of PO and PJ fit over time. Before introducing LMX as an enabling condition, we lay out the general rationale for the spillover between PO fit and PJ fit.

High PO fit implies that the employee and the organization have shared standards concerning what is important and “a common frame for describing, classifying, and interpreting events”, such that the employee–organization relationship is characterized by high exchange of information, improved communication, low likelihood of misunderstandings, high predictability and high levels of trust (Edwards and Cable, 2009: 656). This is likely to facilitate both the match between supplies of the job and employee needs (NS fit), and the match between job demands and employee abilities (DA fit).

Employees who report high PO fit share standards with the organization concerning what is important, and they have similar motives and goals, which help them to predict what will occur in the organization (Edwards and Cable, 2009). These characteristics enhance the likelihood that, for an employee with high PO fit, the job meets the employee’s expectations, and also offers the pay, benefits, training, content in terms of interesting and challenging work, promotion opportunities, recognition and decision-making latitude to fulfill the employee’s needs. Supporting this line of reasoning, Kristof (1996) suggests that individual and organizational values may be framed as the mechanisms underlying employee desires and the rewards entailed in employment relationships. In other words, what is important in an organization underpins the types of rewards that the organization supplies to employees in certain jobs (Schein, 1992), and what is important to an individual underpins his or her desires (Hogan, 1991). Thus, value congruence between an employee and her organization (PO fit) may influence the employee’s perceptions of the extent to which her job fulfills her needs.

PO fit may also affect DA fit, on the basis of the assumption that individuals develop abilities that enable them to pursue what they value – that is, they put effort into acquiring competencies in the context of specific professional interests as well as broader life interests (e.g. Tharenou, 2003), and that job demands often reflect what the organization considers important (e.g. Schein, 2006). Furthermore, individuals who experience higher PO fit are more likely to understand organizational as well as job-specific rules, norms and performance expectations, and they perceive less role ambiguity (Edwards and Cable, 2009). As a result, they have a better understanding of which abilities they need to possess and further develop (e.g. with additional training) in order to meet their job demands. Also, over time, they are more likely to be positively reinforced for performing particular tasks well and to be removed from roles that do not suit their abilities (Kristof-Brown et al., 2002).

The moderating role of leader–member exchange

As noted above, limited understanding of the conditions under which fit perceptions develop over time is a significant lacuna in the fit literature (Kristof-Brown et al., 2002, 2005). In line with the context-based approach to management theory (Bamberger, 2008; Johns, 2006), which suggests that the broader work setting within which employees are embedded has important implications for their fit perceptions and for the consequences of these perceptions over time (Shipp and Jansen, 2011), we propose that contextual factors may influence how PO and PJ fit perceptions evolve to affect each other as well as specific outcomes over time. In particular, as managers largely serve as intermediaries in the relationship between employees and the organization – framing the context in which employees work, which is essential in explaining employee perceptions and behavior (Aselage and Eisenberger, 2003) – we suggest that the nature of employee relationships with their managers may be relevant for understanding the development of fit perceptions (Kristof-Brown et al., 2005).

Employees’ relationships with their supervisors have been often conceptualized in terms of quality of LMX (Graen and Uhl-Bien, 1995). LMX is premised on notions of social exchange and reciprocity (Blau, 1964; Gouldner, 1960). LMX theory suggests that a leader develops a unique exchange relationship with each follower, and that the quality of this relationship affects the subordinate’s attitudes and behaviors (Gerstner and Day, 1997; Graen and Uhl-Bien, 1995; Liden and Maslyn, 1998). High-quality LMX relationships are characterized by high levels of trust, professional respect, information sharing, career development opportunities, support, loyalty, and formal and informal rewards. Furthermore, followers in higher-quality LMX relationships often enjoy more challenging assignments, sponsorship and greater access to information relevant to the job. In contrast, a low-quality LMX relationship is marked by a lack of respect, trust and resource sharing (e.g. Davis and Gardner, 2004; Ilies et al., 2007; Liden et al., 1997).

The benefits associated with high-quality LMX (Erdogan et al., 2004; Liden et al., 1997) are likely to strengthen the relationship between PO fit and PJ fit perceptions over time, because employees enjoying these benefits are likely to more easily translate PO fit perceptions into PJ fit perceptions. As mentioned above, high PO fit is characterized by high-quality communication and high exchange of information among organizational members. Employees with high PO fit share a common frame of understanding with the organization, are likely to collaborate well with colleagues, and are expected to have a good understanding of job expectations and desired behaviors, all of which are likely to enhance need fulfillment and work adjustment. Having a high-quality LMX relationship, characterized by high levels of trust and support, may further enhance an employee’s understanding of the firm’s values and expectations, as well as further encourage cooperation and communication (e.g. with colleagues). Thus, high LMX may contribute to employees’ ability to leverage the benefits of high PO fit (Erdogan et al., 2004; Settoon et al., 1996), as high LMX provides employees with better access to the resources they need to fulfill their needs and meet their job demands.

In this respect, prior research suggests that while the organization is often perceived as an abstract entity, employees experience the workplace very directly through their relationship with their supervisor. As a result, the immediate supervisor may have a role in facilitating the more distant exchange relationships that employees maintain with the organization (e.g. Wayne et al., 1997). As such, the general shared understandings that are at the heart of PO fit perceptions are likely to be accelerated under conditions of high LMX – further enabling PJ fit – because LMX provides an immediate, direct channel for aid and know-how between employees and their supervisors, and thus, the organization as a whole. This should enable employees to more accurately frame their job expectations and also to better equip themselves with the knowledge and skills necessary for the job.

Finally, employees with higher-quality LMX relationships often perceive themselves as having greater impact on their environment and control over their own actions (Erdogan et al., 2004), which is likely to make them better able to use their high PO fit to meet the demands of the job, and to better align with their personal needs. Taken together, these arguments suggest that, among employees experiencing high LMX, the relationship between PO fit and PJ fit over time is likely to be amplified.

At the same time, when LMX is low, we expect the relationship between PO fit and PJ fit over time to be attenuated. The reasoning is that employees experiencing poor LMX relationships are less likely to enjoy the benefits of high trust and support from their supervisor, are less likely to engage in resource sharing, and are more likely to perceive themselves as having less control over their actions. These characteristics provide low-LMX employees with fewer opportunities to increase their PJ fit when they experience high PO fit. Thus, although they understand expectations and desired behaviors (because of high PO fit), they have less access to resources that might enable them to fulfill their needs and meet the job demands (because of low LMX). Thus we propose:

Hypothesis 1: LMX at T1 moderates the positive relationship between PO fit at T1 and PJ fit at T2, such that the relationship between PO fit and PJ fit is amplified when LMX quality is high.

Subsequent effects on turnover

Employee turnover is a critical problem for many organizations, as it imposes high costs on both organizations and employees. For organizations, turnover involves direct costs (payments to staff members who work overtime to cover for departing employees, costs of recruitment, selection and training of new staff) and indirect costs (in the form of lost knowledge and reduced productivity). The departing employees incur costs associated with job search, unemployment, learning and adjusting to a new job, and so forth (e.g. Holtom et al., 2008). Research suggests that the employee’s turnover decision is influenced by fit perceptions, with lack of fit seriously undermining employees’ intentions to retain their membership with the organization. As suggested above, TWA serves as a framework within which to predict the outcomes of the balance between the individual and the work environment, and in particular – tenure, that is, remaining on the job (the predictive model; Cable and DeRue, 2002; Cable and Judge, 1996; Dahling and Librizzi, 2015; Dawis and Lofquist, 1984; Hoffman and Woehr, 2006; Kristof-Brown et al., 2005; O’Reilly et al., 1991; Tak, 2011).

We incorporate the relational context to propose that LMX may influence not only the degree to which PO fit affects PJ fit (as argued above), but also the degree to which PJ fit affects turnover. The benefits inherent in LMX may divert employees’ attention from the immediate, technical aspects of their specific jobs and the question of whether or not they are well suited to these jobs, and instead lead them to focus on general, broader aspects of their work environment that are informed by their relationship with their supervisor. Subordinates in high-quality LMX relationships are therefore likely to reciprocate by exhibiting strong commitment, loyalty and trust (e.g. Settoon et al., 1996), even if they are experiencing PJ misfit. That is, for employees with higher-quality LMX, PJ fit may factor less into the work experience and therefore be less influential in affecting turnover. At the same time, under conditions of low-quality LMX, employees may look at other parts of the work environment to increase the quality of their work experience. One such alternative focus may be PJ fit, which reflects the extent to which employees are making proper use of their abilities and addressing their professional needs. Employees for whom these needs are more salient may be more likely to leave the organization if they are not being satisfactorily fulfilled (poor PJ fit). In sum, we propose the following:

Hypothesis 2: LMX moderates the relationship between PJ fit at T2 and turnover at T3, such that the relationship between PJ fit and turnover is weaker when LMX quality is higher.

Method

In November 2009 (T1) and November 2010 (T2), employees of 147 teams in 19 locations of an elderly care organization in the Netherlands were asked to fill out a questionnaire, which took 15 minutes to complete. Questionnaires were distributed via internal mail, accompanied by a letter from the HR director that encouraged employees to participate; the letter explained that participation was voluntary and that responses would be kept confidential. Stamped return envelopes were provided, by which completed questionnaires were sent directly to the authors. After three weeks, the HR director sent a reminder to all employees in order to further encourage participation in the survey.

The target sample at both T1 and T2 comprised all individuals employed in the firm at the time of survey administration. At T1, a questionnaire was sent to all 1629 employees, with 487 employees providing completed questionnaires (a response rate of 30%). At T2, all 1862 employees received a questionnaire, and 550 employees provided completed questionnaires (a response rate of 30%). Because of missing data that varied somewhat between variables, the effective sample size ranged between 160 and 166 persons who responded to all questions at both T1 and T2. We used all available data in the analyses (Kline, 2005). These respondents reported to 82 supervisors (on average 1.95–2 respondents per supervisor). Two years after T2, in December 2012, we received archival turnover data regarding all participating employees. During this two-year period, 7% of the respondents had left the organization. This figure is relatively low compared with figures reported in several studies among different OECD countries, where turnover rates ranged between 9% and 15% per year (Simoens et al., 2005). It is consistent, however, with figures reported by Statistics Netherlands, which indicate that average turnover rates in nursing and care homes, ranging between 2.5% and 5.5% per year, are consistently below overall national average turnover rates, because of government measures to cut entitlements under the Exceptional Medical Expenses Act (Bakker et al., 2010).

The sample included nurses, therapists and physicians (85%), in addition to support staff. Respondents’ mean age was 43 years (SD = 11.37), and their average tenure was 12 years (SD = 9.28). Ninety percent were female. In terms of education, 47% of the respondents had completed a vocational education program, 16% held a bachelor’s or higher degree, and 37% were high school graduates. Consistent with Statistics Netherlands’ figures on the average working hours of nurses in the health care sector in the Netherlands, 35% of the respondents worked 2.5 days per week or fewer, and 31% worked four days per week or more.

Measures

Unless otherwise stated, employees responded to survey items using a seven-point Likert-type scale, ranging from 1 = “strongly disagree” to 7 = “strongly agree”. The survey was administered in Dutch. We used the translation-back-translation procedure (Brislin, 1980) to translate all items.

PO fit (T1 and T2) was assessed using Cable and DeRue’s (2002) three-item measure, including “The things I value in life are very similar to the things that my organization values.”

PJ fit (T1 and T2) was measured using Cable and DeRue’s (2002) three-item scale for DA fit (e.g. “My abilities and training are a good fit with the requirements of my job”), and three-item scale for NS fit (e.g. “The attributes that I look for in a job are fulfilled very well by my present job”).

LMX quality (T1) was measured by the LMX-7 (Graen and Uhl-Bien, 1995). A sample item: “My supervisor recognizes my potential.”

Actual turnover data, per December 2012, were provided by the organization. For each employee, we coded this variable as 0 if the employee remained with the organization between T2 and T3 and 1 if the employee left during this time.

Control variables

To rule out potentially spurious relations, in all our analyses we controlled for age (in years), gender (1 = male, 2 = female), tenure (in months), and education level. These control variables are commonly used in studies on fit and turnover (Griffeth et al., 2000; Mossholder et al., 2005).

Measurement model

The measurement model had seven latent factors (PO fit T1 and T2, DA fit T1 and T2, NS fit T1 and T2, and LMX) and provided an acceptable fit to the data (χ² = 431.76 (d.f. = 254, N = 190), p < .01, Comparative Fit Index (CFI) =.96; Tucker Lewis Index (TLI) = .95, Root Mean Square Error of Approximation (RMSEA) = .06, Standardized Root Mean Square Residual (SRMR) = .05). To test for measurement invariance of the three fit scales, we estimated one model in which factor loadings relating the indicators to their respective time factors were freely estimated across time periods; the fit of this model was acceptable (χ² = 214.25 (d.f. = 120, N = 190), p < .01, CFI =.97; TLI = .96, RMSEA = .06, SRMR = .05). A more constrained model in which respective factor loadings were set to be equal across T1 and T2 fit the data just as well as the less-constrained measurement model (χ² = 217.57 (d.f. = 126, N = 190), p < .01, CFI =.97; TLI = .97, RMSEA = .06, SRMR = .05), indicating adequate discriminate validity of the scales across time.

Analysis strategy

The data we collected corresponded to three levels of analysis: individual, team (supervisor) and location. Therefore, we used multilevel logistic regression path analysis in Mplus to test our hypotheses. Intra-class correlation coefficient (ICC) values indicated that there is little variance of the variables in our model to be explained at the team level (ICC values of DA fit, NS fit and PO fit at T2, and Turnover are 0). This suggests that team-level analyses are unnecessary and that relationships in our model should be viewed at the individual level of analysis. Also, researchers stress that modeling paths in complex multilevel models should be guided by theoretical choices (e.g. Hofmann, 1997). Therefore, as our theoretical focus was on variation across individuals on PO fit, PJ fit, LMX and actual turnover, we modeled these variables at the individual level and controlled for variance on the team and location levels (levels 2 and 3). Before we performed the analyses, we checked the data for outliers, and normality of the variables. For missing data we used maximum-likelihood estimation. This method partitions the cases into subsets with the same patterns of missing observations. All available information is extracted from each subset, and all cases are retained in the analysis. Therefore, much statistical power is restored, and bias due to missing data is eliminated to a large extent. Studies have shown that this method generally outperforms other methods to deal with missing data (Collins, 2006; Graham, 2009; Kline, 2005).

Path analysis was conducted to test both Hypotheses 1 and 2 in a single model, that is, the moderation effect of LMX and subsequent effects on turnover. Our dependent variable in this model was a binary variable. Therefore, we performed multilevel logistic regression path analysis. To calculate model fit for the multilevel logistic regression path analysis, we used log-likelihood ratios and the Akaike Information Criterion (AIC). The AIC takes into account both the log likelihood and the number of parameters in the model, with lower AIC values indicating better fit to the data. We compared the fit of different models reflecting the hypothesized relationships between PO and PJ fit, with and without controlling for reversed relationships, and we used the results obtained from the best-fitting model. Also, we controlled for the level of the respective type of fit at T1.

Results

Descriptive statistics, reliabilities and correlations are provided in Table 1.

Table 1.

Means, standard deviations and correlations.

Variable	M	SD	1	2	3	4	5	6	7	8	9	10	11
1. Age	44.54	11.41
2. Gender (1 = male, 2 = female)			−.10
3. Tenure (in months)	152.41	115.45	.49**	−.04
4. Education level			−.02	−.21**	.02
5. Leader–Member Exchange T1	2.93	.68	.04	.08	−.03	−.10	(.89)
6. Person–Organization fit T1	5.13	1.04	.04	.13	−.10	−.00	.39**	(.89)
7. Person–Organization fit T2	5.16	1.03	.09	.17*	−.06	.04	.35**	.46**	(.92)
8. Demands–Abilities fit T1	5.62	.95	.11	.04	.01	.09	.30**	.37**	.11	(.90)
9. Demands–Abilities fit T2	5.55	1.08	.22**	.12	.17*	.05	.33**	.31**	.47**	.44**	(.91)
10. Needs–Supplies fit T1	5.26	1.16	.23**	.02	.00	−.07	.37**	.45**	.24**	.66**	.40**	(.95)
11. Needs–Supplies fit T2	5.23	1.23	.22**	.14	.10	−.03	.38**	.35*	.49**	.40**	.72**	.62**	(.96)
12. Actual turnover T3 (0 = staying, 1 = leaving)			.06	−.38**	.02	.11	−.06	−.03	−.07	−.07	−.15	−.03	−.15

M = mean; SD = standard deviation; N = 160. Alphas are in parentheses. *p < .05, **p < .01.

Hypothesis 1 stated that LMX moderates the relationship between PO fit at T1 and PJ fit at T2, and Hypothesis 2 proposed that LMX moderates the relationship between PJ fit at T2 and turnover at T3. We tested these hypotheses simultaneously in one model. Before the proposed model was tested, the independent and moderator variables were grand-mean centered, and interaction variables were calculated using the product of the centered variables. We tested a model that included the relationships between PO fit and each dimension of PJ fit, controlling for reversed relationships between PO and PJ fit (third column in Table 2), as well as for PO and PJ fit at the previous point in time. The results show that the interaction term of PO fit at T1 and LMX is significantly and positively related to both DA fit (β = .28, p < .01) and NS fit (β = .21, p < .01) at T2 (see Table 2). These results support Hypothesis 1. Figures 2a and 2b, respectively, show these relationships, plotted for LMX values corresponding to one standard deviation above and below the mean of LMX. Simple slopes analyses show that the slopes for low-quality LMX are non-significant, and that the slopes for high-quality LMX are positive and significant (PO fit T1 – DA fit T2: .36, p < .01; PO fit T1 – NS fit T2: .23, p < .01).

Table 2.

Results of multilevel logistic regression path analysis testing moderated mediation.

	DA fit T2	NS fit T2	PO fit T2	Turnover
Age	.12* (.05)	.07 (.06)	.10 (.10)	−.01 (.17)
Gender (1 = male, 2 = female)	.09 (.09)	.12* (.05)	.12 (.08)	−.36** (.09)
Tenure (in months)	.10 (.05)	.06 (.05)	−.07 (.09)	.02 (.15)
Education level	.08 (.06)	.07 (.06)	.13 (.09)	.11 (.21)
Demands–Abilities fit T1	.18 (.12)	−.09 (.06)	−.19* (.08)	−.22 (.17)
Needs–Supplies fit T1	.09 (.10)	.56** (.08)	.11 (.11)	.43 (.27)
Person–Organization fit T1	.16 (.09)	.08 (.06)	.39** (.07)	−.03 (.21)
Leader–member exchange T1	.26** (.08)	.23** (.06)	.24* (.10)	.04 (.15)
PO fit T1 × LMX	.28** (.07)	.21** (.07)	.05 (.06)
DA fit T1 × LMX	−.13 (.10)	−.07 (.10)	−.05 (.10)
NS fit T1 × LMX	−.02 (.09)	.06 (.13)	.16* (.07)
Demands–Abilities fit T2				.27 (.17)
Needs–Supplies fit T2				−.53** (.17)
Person–Organization fit T2				.03 (.19)
PO fit T2 × LMX				−.17 (.11)
DA fit T2 × LMX				.45** (.14)
NS fit T2 × LMX				−.19 (.17)
R²	R² = .36**	R² = .48**	R² = .31**	R² = .52**
Log likelihood = −577.99AIC = 1257.98

N = 162. Standardized coefficients are presented. Standard errors are presented between parentheses. DA = Demands–Abilities; T = Time; NS = Needs–Supplies; PO = Person–Organization; LMX = Leader–Member Exchange.

p < .05 **p < .01.

Figure 2a.

The moderating role of LMX in the relationship between PO Fit at T1 and DA fit at T2.

Figure 2b.

The moderating role of LMX in the relationship between PO Fit at T1 and NS fit at T2.

Hypothesis 2 proposed that LMX moderates the relationship between PJ fit at T2 and turnover at T3. Table 2 shows that, contrary to our expectations for an attenuation effect of LMX, the coefficient for the interaction between DA fit at T2 and turnover is significant and positive (β = .45, p < .01), suggesting an amplification effect. Results were plotted for LMX values corresponding to one standard deviation above and below the mean (Figure 3). Simple slopes analyses show that the slope for low-quality LMX is non-significant, and the slope for high-quality LMX is positive and significant (1.86, p < .05). LMX had no moderating role in the relationship between NS fit at T2 and turnover. Instead, NS fit at T2 was directly and strongly related to turnover (β = –.53, p < .01). Thus these findings provide no support for Hypothesis 2.

Figure 3.

The moderating role of LMX in the relationship between DA fit at T2 and turnover at T3.

Discussion

The current study focused on the moderating role of LMX in (1) the relationships between PO and PJ fit perceptions over time, and (2) the subsequent effects of PJ fit perceptions on turnover. Results suggest that PO fit at T1 is positively related to both types of PJ fit (DA fit and NS fit) at T2, only for employees in high-quality LMX relationships. For employees in low-quality LMX relationships, we found no significant association between PO fit and PJ fit.

Our results thus suggest that LMX quality is highly relevant for understanding the relationship between fit dimensions as time passes. This points to the salience of the supervisor as a main source of influence on fit perceptions and their interpretation. The support, trust, cooperation and resource sharing associated with a high-quality LMX relationship seem to be necessary for PO fit perceptions to affect PJ fit perceptions over time. A closer look at Figures 2a and 2b reveals that the left side of each figure (i.e. low PO fit) shows no difference in the level of PJ fit between low and high LMX. Yet the right side of each figure (i.e. high PO fit) does show the expected difference in the level of PJ fit between low and high LMX. Thus, high LMX is only effective in enhancing PJ fit perceptions when employees perceive high PO fit – a combination of high PO fit at T1 and a high-quality LMX relationship with one’s supervisor is necessary in order to develop high PJ fit perceptions over time. Our findings support the TWA (Dawis and Lofquist, 1984), which regards work adjustment as a continuous process in which employees seek to achieve and maintain correspondence with their work environment. We extend this idea by showing that LMX quality is an important enabler of – or may even be an important requirement for – this process of work adjustment to occur.

In looking at the reversed relationships between PO and PJ fit (included as control), an interesting finding emerged: the NS dimension of PJ fit at T1 was related to PO fit at T2, but also only when LMX was high (β = .16, p < .05; here too, the slope is insignificant for low-quality LMX, yet it is positive and significant for high-quality LMX: β = .24, p < .05). This may suggest that, under conditions of high LMX, the relationship between PO and PJ fit perceptions, or supplementary and complementary fit, is bidirectional rather than unidirectional. This means that PO fit and PJ fit influence each other over time, and, in line with suggestions of Jansen and Kristof-Brown (2006), spillover effects seem to occur from one fit domain to the other.

Another aim of this study was to examine effects of PJ fit on turnover, measured two years after T2. As only a few studies have examined the role of contextual moderators – factors that frame employee judgment and decision making – in the relationship between fit and employee behaviors (Kristof-Brown et al., 2005), we sought to take account of LMX in the relationship between PJ fit perceptions at T2 and turnover. We expected a compensatory effect, such that for employees with high LMX, PJ fit would be less influential in lowering turnover. However, contrary to our expectations, we found that for low-quality LMX there was no significant relationship between DA fit perceptions and turnover, whereas for high-quality LMX, DA fit enhanced rather than decreased turnover. Thus, the combination of having higher DA fit and a high-quality LMX relationship was associated with higher rather than lower turnover. This finding could be explained using the concept of self-efficacy (Bandura, 1986). Greguras and Diefendorff (2009) argue that DA fit is similar to self-efficacy (Bandura, 1986). Self-efficacy refers to people’s belief about their ability to accomplish a task (Bandura, 1986), which is close to the belief that one has the skills and abilities to perform well. People who perceive their abilities as closely matching the requirements of their environment (i.e. high DA fit) may therefore be more motivated to master challenges, focus less on their deficiencies (Greguras and Diefendorff, 2009), and have higher self-esteem (Wang et al., 2011). In line with this, research has shown that people with high perceptions of DA fit feel more competent in performing their jobs (Greguras and Diefendorff, 2009). This may imply that, owing to their increased mastery and confidence, employees with high DA fit perceptions may view the organization as offering limited opportunities and challenges for them. When these employees have high-quality LMX relationships with their supervisors, they experience high levels of trust, support, professional respect and development opportunities (Davis and Gardner, 2004; Ilies et al., 2007; Liden et al., 1997), such that they may be more confident in their ability to seek other job opportunities outside the organization, resulting in higher turnover.

A somewhat related explanation for this finding may be that employees whose survey results indicated high DA fit may in fact have felt overqualified (having more abilities – education, experience and training – than a job requires; Khan and Morrow, 1991). This, together with a supportive supervisor who may be encouraging them to seek employment opportunities outside the firm (high LMX), may indicate misfit rather than fit. Thus, and given that overqualification is typically measured in terms of employee perceptions of mismatch between one’s abilities and the requirements of one’s job (Maynard et al., 2006), the DA fit sub-scale may have been capturing overqualification, at least to some extent. A recent study by Liu et al. (2015) provides some initial support for this notion. More specifically, these authors conceptualized overqualification as a type of poor PJ fit, and suggested that “in processing their person–job misfit, overqualified employees might cognitively appraise themselves as less worthy organizational members” (p. 250). Their findings indicate that overqualification is positively related to counter-productive work behaviors. Our study suggests that a supportive work environment (high LMX), instead of buffering against such negative outcomes, may create a sense in employees of greater tolerance on the part of their supervisors, and thus further encourage withdrawal.

Cultural factors may underlie the finding that low DA fit–low LMX individuals are less likely to quit. Compared with firms in other western countries, Dutch firms may be less likely to subject employees to performance-related pressures. For example, US firms are more likely to provide performance-based rewards than are Dutch firms. Moreover, incentives in Dutch firms are more likely to be based on nonfinancial performance measures (Merchant et al., 2011). These approaches are also reflected in LMX relations; US managers, relative to their Dutch counterparts, more strongly believe in the power of incentives (Jansen et al., 2009). These differences in management philosophies between the two countries, together with Lawrence’s (1991) observation that Dutch managers are tolerant in that they are unlikely to dismiss poor performers, may help explain our finding that low-LMX individuals experiencing lack of fit showed low turnover.

We found that perceived NS fit at T2 was negatively associated with turnover, regardless of the level of LMX, showing that perceived NS fit seems to be an important predictor of turnover. This finding is in line with Cable and DeRue (2002), who have argued that NS fit may be the most important type of fit from an employee perspective. Taken together, the results confirm that, in line with previous studies, DA fit and NS fit are likely to serve as salient cues used as input for work-related decisions (Resick et al., 2007), and that the relative importance of PJ fit in predicting turnover decisions seems to increase over time (Kristof-Brown et al., 2002), such that PJ fit perceptions, rather than PO fit perceptions, ultimately affect turnover decisions.

Looking at the full model (see Figure 1), our findings suggest a process in which PO fit perceptions affect changes in PJ fit perceptions over time and influence subsequent turnover – but only for employees in high-quality LMX relationships. This observation may contribute to research on the attraction-selection-attrition (ASA) framework (Schneider et al., 1995), which was originally developed to explain organizational homogeneity, and which suggests that people are differentially attracted to, and are likely to stay in, organizations that fit with their values and interests. In particular, Kristof-Brown et al. (2005) have suggested that, although researchers frequently use the ASA framework to investigate individual fit in organizations, the framework cannot be reliably generalized to the individual domain without further investigation of factors that influence changes in individuals’ levels of fit over time, from pre-entry to turnover.

Our findings extend the ASA framework in two ways. First, our results show that for individuals in high-quality LMX relationships, PO fit perceptions affect turnover via enhancing PJ fit perceptions. This observation suggests that, even after employees achieve an initial state of fit, their turnover decisions are influenced by their subsequent perceptions of fit. Notably, our results suggest that the ASA cycle may not function in a similar way for all employees; LMX influences the extent to which different types of fit affect each other, and subsequent turnover decisions. Second, our results challenge the assumption that employees who do not fit leave the organization. In fact, results show that higher perceived NS fit is associated with lower turnover, but that higher perceived DA fit is associated with higher turnover among employees who are in high-quality LMX relationships. More research using longitudinal data is needed in order to confirm the process suggested by our results.

Another surprising finding was the difference in results for the two dimensions of PJ fit: DA fit and NS fit. Most studies to date have either combined both dimensions into one construct of PJ fit or complementary fit (e.g. Tak, 2011) or assumed that both dimensions have similar effects on various outcomes of interest (e.g. Wang et al., 2011). Our results for NS fit were in line with previous studies. However, higher perceived DA fit at T2 was associated with higher rather than lower turnover among employees in high-quality LMX relationships. This suggests that, among employees with high-quality LMX, PO fit might influence turnover decisions through counteracting mechanisms: PO fit lowers turnover through enhancing NS fit, but enhances turnover by enhancing DA fit. Whereas the effects of NS fit are likely to be based in the need fulfillment argument proposed above, self-efficacy may play a role in explaining the effects of DA fit (DeRue and Morgeson, 2007). More specifically, because self-efficacy impacts how individuals interpret the meaning of and attributions about their work performance (Bandura, 1986), it may be that self-efficacy, embedded in high perceived DA fit (i.e. individuals attribute their performance to their own ability), may have a strong enough effect such that PO fit becomes less relevant and quitting one’s job seems an attractive option. In other words, the attributional tendencies associated with high DA fit perception via high self-efficacy may have influenced the nature of the relationship between this component of PJ fit and turnover.

Limitations and suggestions for future research

The key strengths of our study include the two-wave dataset and subsequent objective turnover data. However, this study has several limitations, some of which may offer opportunities for future research. First, we conducted our research in a health care organization, a specific context that may have affected our results. In health care, where employees are predominately female (as clearly evident in our sample comprising of 90% women), most jobs require a high degree of cooperation and teamwork and frequent contact with patients or clients. Therefore, for employees in this industry, relationships with supervisors, colleagues and clients may be more salient compared with their perceptions of the organization as a whole. Future research could replicate our study in other sectors (and with different gender compositions) to see whether the relationships hold. Researchers may also consider person–team fit and customer-related variables as potential predictors of withdrawal in health care as well as other service contexts.

In addition, concepts other than LMX have been offered to capture employees’ relationships with their supervisors. One such example is the concept of transformational leadership, referring to leaders who motivate followers beyond immediate self-interests via charismatic influence, inspiration, intellectual stimulation or individualized consideration (e.g. Bass and Avolio, 1995). In particular, Hoffman et al. (2011) found that group-level PO fit mediated the effect of transformational leadership on work group effectiveness. Future research may thus consider transformational leadership as well as other related factors, such as supervisor support, as potential moderators.

Recent definitions of longitudinal research have emphasized the use of three or more waves of data collection (e.g. Ployhart and Vandenberg, 2010), suggesting that the capacity of two-wave prospective studies to adequately assess change may be somewhat limited. Future research may benefit from data collected in three or more time points. In addition, LMX was measured only at T1, so we could not take changes in LMX between T1 and T2 into account. Future research may include LMX at multiple time points in order to account for the effects of changes in LMX. In addition, within each time period, our data may be subject to common source bias, which may have led to an overestimation of relationships between variables within each time period.

The two-year turnover lag used in this study was relatively long, which may have influenced our results. However, meta-analytic evidence (Griffeth et al., 2000) has shown that the association between employee attitudes and turnover is weaker when a longer time lag is used. Thus, if our chosen time lag has influenced our results, results are likely to represent an underestimation rather than overestimation of the relationship between fit and turnover. Still, to rule out other factors that potentially influence turnover, we urge researchers investigating fit as a dynamic construct to have a shorter time separation between measurements of fit and turnover.

As discussed above, the turnover rate in our sample, though consistent with figures reported by Statistics Netherlands for nursing and care homes, was low. Indeed, several turnover studies have encountered concerns over low quit rates, and these concerns have motivated researchers to adopt long measurement windows to allow for more leavers to be included in their analysis. Yet Griffeth et al. (2000) concluded from their meta-analysis that the “beneficial effects of expanded termination rates on improved predictive accuracy may be overstated” (p. 484). According to their findings, turnover rate did not influence any of the predictor-quit relationships. Future research could still replicate our results in a setting in which turnover rate is higher.

Finally, the actual turnover data we used did not distinguish between voluntary and involuntary turnover. Our theoretical framework focuses on voluntary turnover, and information we received from the organization indicates that involuntary turnover (i.e. dismissal) is rare, which makes it likely that most instances of turnover in our data were indeed voluntary. Notably, in light of prior research suggesting that effects on turnover may be even stronger if the data only contain voluntary turnover (e.g. Koys, 2001), we assume that, if anything, our results err towards conservative estimates. Nonetheless, future research may distinguish between voluntary and involuntary turnover.

Conclusions and implications

Our study suggests that PO fit perceptions affect PJ fit perceptions over time, only when LMX quality is high. More generally, our findings suggest that research should take into account conditions under which different types of fit perceptions develop and influence one another over time. Future research might consider other contextual moderators, in addition to LMX, that could influence perceptions of fit. In addition, differences identified between the two dimensions of PJ fit – DA fit and NS fit – may have implications for theory and should be further explored.

Our results also have some practical implications. The finding that spillover effects of one type of fit on the other only occur when LMX is high suggests that supervisors may have the most potent role in the interoperation and development of fit perceptions. Organizations could focus on developing high-quality relationships between supervisors and subordinates to facilitate the development of employees’ fit perceptions over time. However, organizations should recognize that, among employees who perceive their abilities as matching the demands of the job (high DA fit), a high-quality relationship with one’s supervisor can provide motivation to leave the company and seek challenges elsewhere. In fact, in these cases, the supervisor may even be a driving force for moving on to a better position, not necessarily with the same employer. This possibility further suggests that firms might seek to differentiate between DA misfit resulting from excessive demands and DA misfit resulting from excessive abilities (i.e. overqualification), with the former incorrectly interpreted as DA fit, and potentially – particularly under conditions of high LMX – encouraging employees to leave. Finally, fulfilling the needs of employees such that they perceive high NS fit seems to be the most salient factor in explaining turnover decisions. Organizations may aim to enhance their employees’ perceptions of NS fit in order to keep turnover low.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Corine Boon is Assistant Professor of Human Resource Management at the University of Amsterdam Business School, the Netherlands. She received her PhD from the Erasmus University Rotterdam, the Netherlands. Her research focuses on strategic HRM, the HRM implementation process, HRM and leadership, and person–environment fit. Her research has been published in journals including Journal of Management, Journal of Personnel Psychology, Journal of Managerial Psychology, Human Resource Management, Human Resource Management Journal and the International Journal of Human Resource Management. [Email: c.t.boon@uva.nl]

Michal Biron, Department of Business Administration, Faculty of Management, University of Haifa, Haifa, Israel, and Department of Human Resource Studies, Tilburg University, Tilburg, the Netherlands. Michal is the Chair of the Academy of Management HR Division Ambassadors Program. Her research focuses on the nature of peer relations in the workplace and the implications of such relations for both individuals and organizations, the role of HRM in organizations and employee misbehaviors. Her articles have appeared in Journal of Applied Psychology, Organizational Behavior and Human Decision Processes, Journal of Vocational Behavior, Journal of Occupational Health Psychology, Human Relations and International Journal of Human Resource Management. [Email: mbiron@univ.haifa.ac.il]

References

Aselage

Eisenberger

(2003) Perceived organizational support and psychological contracts: A theoretical integration. Journal of Organizational Behavior 24(5): 491–509.

Bakker

Boon

Riksen

Van Zonneveld

(2010) More turnover and more profit in Dutch health care institutions. Statistics Netherlands. Web magazine, 11 November 2010.

Bamberger

(2008) From the editors: Beyond contextualization: Using context theories to narrow the micro-macro gap in management research. The Academy of Management Journal 51(5): 839–846.

Bandura

(1986) The explanatory and predictive scope of self-efficacy theory. Journal of Social and Clinical Psychology 4(3): 359–373.

Bass

Avolio

(1995) The Multifactor Leadership Questionnaire. Palo Alto, CA: Mind Garden.

Biron

(2010) Negative reciprocity and the association between perceived organizational ethical values and organizational deviance. Human Relations 63(6): 875–897.

Blau

(1964) Exchange and Power in Social Life. New York: Wiley.

Brislin

(1980) Translation and content analysis of oral and written material. In: Triandis

Berry

(eds) Handbook of Cross-Cultural Psychology, Vol.2. Methodology. Boston, MA: Allyn & Bacon. 389–444.

Cable

DeRue

(2002) The convergent and discriminant validity of subjective fit perceptions. Journal of Applied Psychology 87(5): 875–884.

10.

Cable

Edwards

(2004) Complementary and supplementary fit: A theoretical and empirical integration. Journal of Applied Psychology 89(5): 822–834.

11.

Cable

Judge

(1996) Person-organization fit, job choice decisions, and organizational entry. Organizational Behavior and Human Decision Processes 67(3): 294–311.

12.

Cable

Parsons

(2001) Socialization tactics and person-organization fit. Personnel Psychology 54(1): 1–23.

13.

Caldwell

O’Reilly

(1990) Measuring person-job fit with a profile-comparison process. Journal of Applied Psychology 75(6): 648–657.

14.

Collins

(2006) Analysis of longitudinal data: The integration of theoretical model, temporal design, and statistical model. Annual Review of Psychology 57(1): 505–528.

15.

Cooper-Thomas

Van Vianen

Anderson

(2004) Changes in person-organization fit: The impact of socialization tactics on perceived and actual P-O fit. European Journal of Work and Organizational Psychology 13(1): 52–78.

16.

Dahling

Librizzi

(2015) Integrating the theory of work adjustment and attachment theory to predict job turnover intentions. Journal of Career Development 42(3): 215–228.

17.

Davis

Gardner

(2004) Perceptions of politics and organizational cynicism: An attributional and leader–member exchange perspective. The Leadership Quarterly 15(4): 439–465.

18.

Dawis

Lofquist

(1984) A Psychological Theory of Work Adjustment. Minneapolis, MN: University of Minnesota Press.

19.

DeRue

Morgeson

(2007) Stability and change in person-team and person-role fit over time: The effects of growth satisfaction, performance, and general self-efficacy. Journal of Applied Psychology 92(5): 1242–1253.

20.

Edwards

(1991) Person-job fit: A conceptual integration, literature review, and methodological critique. International Review of Industrial and Organizational Psychology 6: 283–357.

21.

Edwards

Cable

(2009) The value of value congruence. Journal of Applied Psychology 94(3): 654–677.

22.

Elfenbein

O’Reilly

(2007) Fitting in: The effects of relational demography and person-culture fit on group process and performance. Group and Organization Management 32(1): 109–142.

23.

Erdogan

Kraimer

Liden

(2004) Work value congruence and intrinsic career success: The compensatory roles of leader-member exchange and perceived organizational support. Personnel Psychology 57(2): 305–332.

24.

Feldman

Vogel

(2009) The aging process and person-environment fit. In: Baugh

Sullivan

(eds) Research in Careers. Charlotte, NC: Information Age Press, 1–25.

25.

Gabriel

Diefendorff

Chandler

. (2014) The dynamic relationships of work affect and job satisfaction with perceptions of fit. Personnel Psychology 67(2): 389–420.

26.

Gerstner

Day

(1997) Meta-analytic review of leader–member exchange theory: Correlates and construct issues. Journal of Applied Psychology 82(6): 827–844.

27.

Gouldner

(1960) The norm of reciprocity: A preliminary statement. American Sociological Review 25(2): 161–178.

28.

Graen

Uhl-Bien

(1995) Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly 6(2): 219–247.

29.

Graham

(2009) Missing data analysis: Making it work in the real world. Annual Review of Psychology 60(1): 549–576.

30.

Greguras

Diefendorff

(2009) Different fits satisfy different needs: Linking person-environment fit to employee commitment and performance using self-determination theory. Journal of Applied Psychology 94(2): 465–477.

31.

Griffeth

Hom

Gaertner

(2000) A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the next millennium. Journal of Management 26(3): 463–488.

32.

Hesketh

(1993) Toward a better adjusted theory of work adjustment. Journal of Vocational Behavior 43(1): 75–83.

33.

Hoffman

Woehr

(2006) A quantitative review of the relationship between person–organization fit and behavioral outcomes. Journal of Vocational Behavior 68(3): 389–399.

34.

Hoffman

Bynum

Piccolo

Sutton (2011) Person-organization value congruence: How transformational leaders influence work group effectiveness. Academy of Management Journal 54(4): 779–796.

35.

Hofmann

(1997) An overview of the logic and rationale of hierarchical linear models. Journal of Management 23(6): 723–744.

36.

Hogan

(1991) Personality and personality assessment. In: Dunnette

Hough

(eds) Handbook of Industrial and Organizational Psychology. Palo Alto, CA: Consulting Psychologists Press, 833–871.

37.

Holtom

Mitchell

Lee

Eberly

(2008) Turnover and retention research: A glance at the past, a closer review of the present, and a venture into the future. The Academy of Management Annals 2(1): 231–274.

38.

Ilies

Nahrgang

Morgeson

(2007) Leader-member exchange and citizenship behaviors: A meta-analysis. Journal of Applied Psychology 92(1): 269–277.

39.

Jansen

Merchant

Van der Stede

(2009) National differences in incentive compensation practices: The differing roles of financial performance measurement in the United States and the Netherlands. Accounting, Organizations and Society 34(1): 58–84.

40.

Jansen

Kristof-Brown

(2006) Toward a multidimensional theory of person-environment fit. Journal of Managerial Issues 18(2): 193–212.

41.

Jansen

Shipp

(2013) A review and agenda for incorporating time in fit research. In: Kristof-Brown

Billsberry

(eds) Organizational Fit: Key Issues and New Directions. Chichester: Wiley-Blackwell, 195–221.

42.

Johns

(2006) The essential impact of context on organizational behavior. The Academy of Management Review 31(2): 386–408.

43.

Khan

Morrow

(1991) Objective and subjective underemployment relationships to job satisfaction. Journal of Business Research 22(3): 211–218.

44.

Kline

(2005) Principles and Practice of Structural Equation Modeling. New York: The Guilford Press.

45.

Koys

(2001) The effects of employee satisfaction, organizational citizenship behavior, and turnover on organizational effectiveness: A unit-level, longitudinal study. Personnel Psychology 54(1): 101–114.

46.

Kristof

(1996) Person-organization fit: An integrative review of its conceptualizations, measurement, and implications. Personnel Psychology 49(1): 1–49.

47.

Kristof-Brown

Jansen

Colbert

(2002) A policy-capturing study of the simultaneous effects of fit with jobs, groups, and organizations. Journal of Applied Psychology 87(5): 985–993.

48.

Kristof-Brown

Zimmerman

Johnson

(2005) Consequences of individuals’ fit at work: A meta-analysis of person-job, person-organization, person-group, and person-supervisor fit. Personnel Psychology 58(2): 281–342.

49.

Lauver

Kristof-Brown

(2001) Distinguishing between employees’ perceptions of person-job and person-organization fit. Journal of Vocational Behavior 59(3): 454–470.

50.

Lawrence

(1991) Management in the Netherlands. Oxford: Clarendon.

51.

Liden

Maslyn

(1998) Multidimensionality of leader–member exchange: An empirical assessment through scale development. Journal of Management 24(1): 43–72.

52.

Liden

Sparrowe

Wayne

(1997) Leader-member exchange theory: The past and potential for the future. In: Martocchio

(ed.) Research in Personnel and Human Resources Management. New York: Elsevier Science/JAI Press, 47–119.

53.

Liu

Luksyte

Zhou

Shi

Wang

(2015) Overqualification and counterproductive work behaviors: Examining a moderated mediation model. Journal of Organizational Behavior 36(2): 250–271.

54.

Locke

(1976) The nature and cuases of job satisfaction. In: Dunnette

(ed.) Handbook of Industrial and Organizational Psychology. Chicago, IL: Rand McNally, 1297–1349.

55.

Maden

Kabasakal

(2014) The simultaneous effects of fit with organizations, jobs and supervisors on major employee outcomes in Turkish banks: Does organizational support matter? The International Journal of Human Resource Management 25(3): 341–366.

56.

Maynard

Joseph

Maynard

(2006) Underemployment, job attitudes, and turnover intentions. Journal of Organizational Behavior 27(4): 509–536.

57.

Merchant

Van der Stede

Lin

(2011) Performance measurement and incentive compensation: An empirical analysis and comparison of Chinese and Western firms’ practices. European Accounting Review 20(4): 639–667.

58.

Mossholder

Settoon

Henagan

(2005) A relational perspective on turnover: Examining structural, attitudinal, and behavioral predictors. Academy of Management Journal 48(4): 607–618.

59.

Muchinsky

Monahan

(1987) What is person-environment congruence? Supplementary versus complementary models of fit. Journal of Vocational Behavior 31(3): 268–277.

60.

O’Reilly

Chatman

Caldwell

(1991) People and organizational culture: A profile comparison approach to assessing person-organization fit. Academy of Management Journal 34(3): 487–516.

61.

Ostroff

Judge

(2007) Perspectives on Organizational Fit. New York: Lawrence Erlbaum.

62.

Park

Monnot

Jacob

Wagner

(2011) Moderators of the relationship between person-job fit and subjective well-being among asian employees. International Journal of Stress Management 18(1): 67–87.

63.

Ployhart

Vandenberg

(2010) Longitudinal research: The theory, design, and analysis of change. Journal of Management 36(1): 94–120.

64.

Resick

Baltes

Shantz

(2007) Person-organization fit and work-related attitudes and decisions: Examining interactive effects with job fit and conscientiousness. Journal of Applied Psychology 92(5): 1446–1455.

65.

Rounds

Dawis

Lofquist

(1987) Measurement of person-environment fit and prediction of satisfaction in the theory of work adjustment. Journal of Vocational Behavior 31(3): 297–318.

66.

Schein

(1992) Organizational Culture and Leadership. San Francisco, CA: Jossey-Bass.

67.

Schein

(2006) Organizational Culture and Leadership. Hoboken, NJ: John Wiley & Sons.

68.

Schneider

Goldstein

Smith

(1995) The ASA framework: An update. Personnel Psychology 48(4): 747–773.

69.

Settoon

Bennett

Liden

(1996) Social exchange in organizations: Perceived organizational support, leader-member exchange, and employee reciprocity. Journal of Applied Psychology 81(3): 219–227.

70.

Shipp

Jansen

(2011) Reinterpreting time in fit theory: Crafting and recrafting narratives of fit in medias res. Academy of Management Review 36(1): 76–101.

71.

Simoens

Villeneuve

Hurst

(2005) Tackling nurse shortages in OECD countries. OECD Health Working Papers 19, OECD Publishing. Available at: https://dx-doi-org.web.bisu.edu.cn/10.1787/172102620474 (accessed 21 February 2016).

72.

Tak

(2011) Relationships between various person-environment fit types and employee withdrawal behavior: A longitudinal study. Journal of Vocational Behavior 78(2): 315–320.

73.

Tharenou

(2003) The initial development of receptivity to working abroad: Self-initiated international work opportunities in young graduate employees. Journal of Occupational and Organizational Psychology 76(4): 489–515.

74.

Tinsley

HEA

(2000) The congruence myth revisited. Journal of Vocational Behavior 56(3): 405–423.

75.

Uhl-Bien

Carsten

(2007) Being ethical when the boss is not. Organizational Dynamics 36(2): 187–201.

76.

Verquer

Beehr

Wagner

(2003) A meta-analysis of relations between person-organization fit and work attitudes. Journal of Vocational Behavior 63(3): 473–489.

77.

Wang

Zhan

Mccune

Truxillo

(2011) Understanding newcomers’ adaptability and work-related outcomes: Testing the mediating roles of perceived P–E fit variables. Personnel Psychology 64(1): 163–189.

78.

Wayne

Shore

Liden

(1997) Perceived organizational support and leader-member exchange: A social exchange perspective. Academy of Management Journal 40(1): 82–111.

79.

KYT

(2009) Affective influences in person-environment fit theory: Exploring the role of affect as both cause and outcome of P-E fit. Journal of Applied Psychology 94(5): 1210–1226.

80.

KYT

(2013) A motivational model of person-environment fit: Psychological motives as drivers of change. In: Kristof-Brown

Billsberry

(eds) Organizational Fit: Key Issues and New Directions. Chichester: Wiley-Blackwell, 21–49.

Temporal issues in person–organization fit,person–job fit and turnover: The role of leader–member exchange

Abstract

Keywords

The temporal relationship between person–organization and person–job fit

The moderating role of leader–member exchange

Subsequent effects on turnover

Method

Measures

Control variables

Measurement model

Analysis strategy

Results

Discussion

Limitations and suggestions for future research

Conclusions and implications

Footnotes

Funding

Notes

References